How to improve Bayesian reasoning: Comment on Gigerenzer and Hoffrage (1995).

G. Gigerenzer and U. Hoffrage (1995) claimed that Bayesian inference problems, which have been notoriously difficult for laypeople to solve using base rates, hit rates, and false-alarm rates, become computationally simpler when information is presented with frequencies based on natural sampling. They made an evolutionary argument for the improved performance. The authors of the present article show that performance can improve with either probabilities or frequencies, depending on the rareness of the events and the type of information presented. When events are rare, probabilities are more difficult to understand than frequencies (i.e., 5 out of 1,000 vs. .005.). Furthermore, when the information is presented as joint and marginal events, nested sets become more apparent. Frequencies based on natural sampling have these desirable properties. The authors agree with Gigerenzer and Hoffrage that frequencies can improve Bayesian reasoning, but they attribute that improvement to the use of mental models that involve elements of nested sets. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Extent and limits of the matching concept in monkeys (Cebus apella).

Two experiments with 11 monkeys investigated the extent and limits of the matching concept. In Exp I, 8 Ss that were experimentally naive with regard to visual stimuli were trained on identity matching with a 2-sample set based on 2-dimensional stimuli. On a subsequent test employing 2 new samples, 4 of the 8 Ss applied the matching rule to the new sample stimuli (as defined by a transfer criterion). Three of these Ss showed substantial savings in learning to match the new samples; 2 of the 3 Ss transferred the matching rule when given a 2nd test with 2 new samples, and the 3rd S showed immediate and complete transfer when tested with a 3rd pair of new stimuli. Results indicate a stronger representation of the matching concept in monkeys than has been demonstrated with pigeons, even in comparable assessment conditions. However, in Exp II, 4 Ss from Exp I failed to transfer the matching rule to steady vs flashing green samples, indicating that the matching concept did not immediately extend beyond the general class of visual stimuli with which it was developed. These and related research findings suggest that representation of the matching concept in animals varies along a specificity–abstractness dimension, reflecting the degree to which the concept is tied to the conditions and context of its development. (45 ref) (PsycINFO Database Record (c) 2011 APA, all rights reserved)     

The psychology of the Monty Hall problem: Discovering psychological mechanisms for solving a tenacious brain teaser.

The Monty Hall problem (or three-door problem) is a famous example of a "cognitive illusion," often used to demonstrate people's resistance and deficiency in dealing with uncertainty. The authors formulated the problem using manipulations in 4 cognitive aspects, namely, natural frequencies, mental models, perspective change, and the less-is-more effect. These manipulations combined led to a significant increase in the proportion of correct answers given by novice participants, largely because of the synergy of frequency-based formulation and perspective change (Experiments 1, 2). In a training study (Experiment 3) frequency formulation and mental models, but not Bayes's rule training, showed significant positive transfer in solving related problems. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Temporally massed CS presentations generate more fear extinction than spaced presentations.

Rodent fear conditioning models both excitatory learning and the pathogenesis of human anxiety, whereas extinction of conditional fear is a paradigm of inhibitory learning and the explicit model for behavior therapy. Many studies support a general learning rule for acquisition: Temporally spaced training is more effective than massed training. The authors asked whether this rule applies to extinction of conditional fear in mice. The authors find that both short- and long-term fear extinction are greater with temporally massed presentations of the conditional stimulus (CS). The data also indicate that once CS presentations are sufficiently massed to initiate, or "induce," extinction learning, further CS presentations are more effective when spaced. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Statistical learning of higher-order temporal structure from visual shape sequences.

In 3 experiments, the authors investigated the ability of observers to extract the probabilities of successive shape co-occurrences during passive viewing. Participants became sensitive to several temporal-order statistics, both rapidly and with no overt task or explicit instructions. Sequences of shapes presented during familiarization were distinguished from novel sequences of familiar shapes, as well as from shape sequences that were seen during familiarization but less frequently than other shape sequences, demonstrating at least the extraction of joint probabilities of 2 consecutive shapes. When joint probabilities did not differ, another higher-order statistic (conditional probability) was automatically computed, thereby allowing participants to predict the temporal order of shapes. Results of a single-shape test documented that lower-order statistics were retained during the extraction of higher-order statistics. These results suggest that observers automatically extract multiple statistics of temporal events that are suitable for efficient associative learning of new temporal features. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Effect of calcium channel blocker nifedipine on anxiety and seizure manifestations of abstinence syndrome in rats after withdrawal of long-term diazepam administration]

PURPOSE: To test whether physician's diagnostic inferences can be improved by communicating information using natural frequencies instead of probabilities. Whereas probabilities and relative frequencies are normalized with respect to disease base rates, natural frequencies are not normalized. METHOD: The authors asked 48 physicians in Munich and Düsseldorf to determine the positive predictive values (PPVs) of four diagnostic tests. Information presented in the four problems appeared either as probabilities (the traditional way) or as natural frequencies. RESULTS: When the information was presented as probabilities, the physicians correctly estimated the PPVs in only 10% of cases. When the same information was presented as natural frequencies, that percentage increased to 46%. CONCLUSION: Representing information in natural frequencies is a fast and effective way of facilitating diagnosis insight, which in turn helps physicians to better communicate risks to patients, and patients to better understand these risks.     

Generalization hypothesis of abstract-concept learning: Learning strategies and related issues in Macaca mulatta, Cebus apella, and Columba livia.

The generalization hypothesis of abstract-concept learning was tested with a meta-analysis of rhesus monkeys (Macaca mulatta), capuchin monkeys (Cebus apella), and pigeons (Columba livia) learning a same/different (S/D) task with expanding training sets. The generalization hypothesis states that as the number of training items increases, generalization from the training pairs will increase and could explain the subjects' accurate novel-stimulus transfer. By contrast, concept learning is learning the relationship between each pair of items; with more training items subjects learn more exemplars of the rule and transfer better. Having to learn the stimulus pairs (the generalization hypothesis) would require more training as the set size increases, whereas learning the concept might require less training because subjects would be learning an abstract rule. The results strongly support concept or rule learning despite severely relaxing the generalization-hypothesis parameters. Thus, generalization was not a factor in the transfer from these experiments, adding to the evidence that these subjects were learning the S/D abstract concept. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Transfer between pose and illumination training in face recognition.

The relationship between pose and illumination learning in face recognition was examined in a yes–no recognition paradigm. The authors assessed whether pose training can transfer to a new illumination or vice versa. Results show that an extensive level of pose training through a face–name association task was able to generalize to a new illumination (Experiments 1 and 3), but an equal level of illumination training failed to generalize to a new pose (Experiment 2). The transfer of pose training was likely to depend on a relatively extensive level of training because the same faces with reduced level of exposure (Experiment 4) were unable to reproduce the transfer effect. The findings suggest that generalization of pose training may be extended to different types of image variation, whereas generalization of illumination training may be confined within the trained type of variation. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Hyperspecificity in Visual Implicit Learning: Learning of Spatial Layout Is Contingent on Item Identity.

Humans conduct visual search faster when the same display is presented for a 2nd time, showing implicit learning of repeated displays. This study examines whether learning of a spatial layout transfers to other layouts that are occupied by items of new shapes or colors. The authors show that spatial context learning is sometimes contingent on item identity. For example, when the training session included some trials with black items and other trials with white items, learning of the spatial layout became specific to the trained color--no transfer was seen when items were in a new color during testing. However, when the training session included only trials in black (or white), learning transferred to displays with a new color. Similar results held when items changed shapes after training. The authors conclude that implicit visual learning is sensitive to trial context and that spatial context learning can be identity contingent. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Exemplar effects in the context of a categorization rule: Featural and holistic influences.

Brooks and colleagues (S. W. Allen & L. R. Brooks, 1991; G. Regehr & L. R. Brooks, 1993) have shown that the classification of transfer stimuli is influenced by their similarity to training stimuli, even when a perfect classification rule is available. It is argued that the original effect obtained by Brooks and colleagues might have resulted from two potential confounding variables. Once these confounds were controlled, the current authors did not replicate Brooks and colleagues' results in Experiment 1. Exemplar effects appeared in Experiment 2 when transfer stimuli were perceptually more similar to training stimuli than in Experiment 1. In Experiment 3, the authors obtained exemplar effects with separated stimuli, a finding that was not predicted by Brooks and colleagues' model. The authors suggest that a close perceptual match between training and transfer stimuli is necessary for the effect to occur, for both integrated and separated stimuli. The nature of this perceptual match, holistic or featural, is discussed. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

New Semantic Learning and Generalization in a Patient With Amnesia.

New learning of semantic information is impaired in amnesia. Several reports have demonstrated that "errorless" learning techniques have allowed patients with amnesia to acquire at least some form of new semantic information, although this information appears to be relatively inflexible. Using insights and principles from connectionist modeling of cortical and medial temporal lobe memory systems, the authors describe why errorless learning procedures act as a poor proxy for the medial temporal lobe, suggesting that they artificially eliminate the variability that defines semantic information. The authors trained a patient with severe amnesia on new semantic sentences both with and without variance and then tested him on both repeated and related novel sentences to assess generalization. He successfully learned new semantic information in both conditions but demonstrated better generalization of semantic concepts following training with variance. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

On the identification of repeatedly presented brief visual stimuli.

Discusses 2 hypotheses concerning the explanation of increasing perceptual performance with temporal repetition of tachistoscopically-presented visual stimuli. R. N. Haber's conception in terms of increasing phenomenal clarity is criticized, primarily for the confounding influence of the "stimulus error." An alternative explanation in terms of the perceptual independence of successive inputs is presented. The assumption of independence is built into a model that combines aspects of C. W. Eriksen's "clearest form" approach with Bayesian decision theory. 4 items of support for independence are presented. Eriksen found (a) no effect of interstimulus interval on accuracy; and (b) no difference in accuracy whether a form was preceded by the same or a different form, or by a flash. Previous studies by the present authors found (a) lack of influence of the locus on an inconsistent stimulus in a sequence, and (b) quite accurate Bayesian predictions of multiple observation performance in 3 experiments using equal prior probabilities of the stimulus states. (39 ref.) (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Can being scared cause tummy aches? Naive theories, ambiguous evidence, and preschoolers' causal inferences.

Causal learning requires integrating constraints provided by domain-specific theories with domain-general statistical learning. In order to investigate the interaction between these factors, the authors presented preschoolers with stories pitting their existing theories against statistical evidence. Each child heard 2 stories in which 2 candidate causes co-occurred with an effect. Evidence was presented in the form: AB→E; CA→E; AD→E; and so forth. In 1 story, all variables came from the same domain; in the other, the recurring candidate cause, A, came from a different domain (A was a psychological cause of a biological effect). After receiving this statistical evidence, children were asked to identify the cause of the effect on a new trial. Consistent with the predictions of a Bayesian model, all children were more likely to identify A as the cause within domains than across domains. Whereas 3.5-year-olds learned only from the within-domain evidence, 4- and 5-year-olds learned from the cross-domain evidence and were able to transfer their new expectations about psychosomatic causality to a novel task. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Shared spatial representations for numbers and space: The reversal of the SNARC and the Simon effects.

In 4 experiments, the authors investigated the reversal of spatial congruency effects when participants concurrently practiced incompatible mapping rules (J. G. Marble & R. W. Proctor, 2000). The authors observed an effect of an explicit spatially incompatible mapping rule on the way numerical information was associated with spatial responses. The authors also observed an effect of an incompatible numerical mapping rule (if smaller than 5, press right; if larger than 5, press left) on the Simon effect. This effect was observed only when both tasks used the same effectors. The results point to a shared spatial representation for explicit spatial information (locations) and implicit spatial information (numbers). (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Training and transfer-of-learning effects in disabled and normal readers: evidence of specific deficits

Two experiments were conducted to assess the specificity of training and transfer deficits in disabled readers, aged 7 to 9 years. Forty-eight children (reading disabled, age-matched normal controls, and reading-level-matched normal controls) participated in both a reading and a nonreading (music) acquisition paradigm. Children received instruction in grapheme-phoneme and symbol-note correspondence patterns, respectively. Posttraining tests (one day and one week) following rule training compared performance on trained exemplar items with performance on untrained transfer items. Results revealed that normal readers were able to transfer their rule knowledge in both the reading and nonreading (music) acquisition paradigms, while disabled readers were proficient only in the music task, and thus demonstrated transfer deficits specific to learning printed language. Transfer was optimally facilitated for all readers when training procedures included not only presentation of exemplars, but also cues for rule derivation and explicit statement of pattern invariances.     

Transfer Effects and Conditional Learning in Rats With Selective Lesions of Medial Septal/Diagonal Band Cholinergic Neurons.

The authors examined visual-spatial conditional learning with automated touchscreen tasks in male Long-Evans rats with selective lesions of medial septal/vertical limb of diagonal band (MS/VDB) cholinergic neurons produced by 192 IgG-saporin. Performance on a conditional task, in which 1 of 2 centrally displayed stimuli directed the rat to respond to an illuminated panel on the left or right, depended on training history: Control rats with experience on other visual tasks performed better than MS/VDB-lesioned rats with similar training histories, whereas this effect was reversed in naive rats. This difference appears to reflect transfer effects present in the control rats that are absent in the MS/VDB-lesioned rats. These findings may suggest that MS/VDB cholinergic neurons play a particular role in the transfer of behavioral experience and flexibility of application of behavioral rules in memory, rather than a role in conditional learning per se. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Bayesian Network for E/M Impedance-Based Damage Identification

A Bayesian network is a probabilistic representation of the multiple cause-effect dependency relationships in a domain. It incorporates human reasoning to deal with sparse data availability and to determine the probabilities of uncertain events. In this paper, a Bayesian network is adopted to model the problem of damage location identification. The damage identification method uses the natural frequency shifts and the undamaged mode shapes of the structure to identify the damage location. The frequency shifts are extracted numerically from a finite-element (FE) model and experimentally from the electromechanical (e/m) admittance signatures of the smart piezoelectric (PZT) transducer bonded to the structure. The undamaged mode shapes are determined from the FE model of the undamaged structure. To incorporate a suitable Bayesian network model, issues of variable selection, variable dependency, probabilistic inference, and error modeling are discussed. The performance of the implemented Bayesian network is verified using both numerical and experimental data. The model is able to accurately determine the damage location, with only a subset of frequency shift data, and eliminated the model errors.     

Discrimination of artificial categories structured by family resemblances: A comparative study in people (Homo sapiens) and pigeons (Columba livia).

Adult humans (Homo sapiens) and pigeons (Columba livia) were trained to discriminate artificial categories that the authors created by mimicking 2 properties of natural categories. One was a family resemblance relationship: The highly variable exemplars, including those that did not have features in common, were structured by a similarity network with the features correlating to one another in each category. The other was a polymorphous rule: No single feature was essential for distinguishing the categories, and all the features overlapped between the categories. Pigeons learned the categories with ease and then showed a prototype effect in accord with the degrees of family resemblance for novel stimuli. Some evidence was also observed for interactive effects of learning of individual exemplars and feature frequencies. Humans had difficulty in learning the categories. The participants who learned the categories generally responded to novel stimuli in an all-or-none fashion on the basis of their acquired classification decision rules. The processes that underlie the classification performances of the 2 species are discussed. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Category Representation for Classification and Feature Inference.

This research's purpose was to contrast the representations resulting from learning of the same categories by either classifying instances or inferring instance features. Prior inference learning research, particularly T. Yamauchi and A. B. Markman (1998), has suggested that feature inference learning fosters prototype representation, whereas classification learning encourages exemplar representation. Experiment 1 supported this hypothesis. Averaged and individual participant data from transfer after inference training were better fit by a prototype than by an exemplar model. However, Experiment 2, with contrasting inference learning conditions, indicated that the prototype model was mimicking a set of label-based bidirectional rules, as determined by the inference learning task demands in Experiment 1. Only the set of rules model accounted for all the inference learning conditions in these experiments. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Delusions and delusional reasoning.

In 2 studies, delusional participants assigned higher probabilities to narratives of actual delusions than participants with no history of delusions; previously delusional participants did not differ significantly from delusional participants or participants with no history of delusions. In Study 2, the authors found that this reasoning bias was specific to delusions and did not generalize to "neutral" text. Familiarity with the content of the delusional narratives played a mediating role in the estimation of their probability, but delusional status also had a significant, independent effect. These findings are consistent with the Bayesian model of delusion formation proposed by D. R. Hemsley and P. A. Garety (1986), and with R. P. Bentall, P. Kinderman, and S. Kaney's (1994) concept of "emotional saliency." A productive area of future research might be to further determine the elements of "emotional saliency" and their impact on the individual steps of the Bayesian model. (PsycINFO Database Record (c) 2010 APA, all rights reserved)